mirror of
https://github.com/NousResearch/hermes-agent.git
synced 2026-07-16 14:32:34 +00:00
a3828a94d0
PR #61587 adds sol-pro/terra-pro/luna-pro to the aggregator lists. Complete those on the native surfaces the same way this PR completes the base tiers: - hermes_cli/models.py: -pro variants in _PROVIDER_MODELS[openai-api]. - agent/usage_pricing.py: alias ("openai", "gpt-5.6-*-pro") onto the base-tier PricingEntry rows — the -pro high-effort modes bill at the SAME per-token rates (verified against OpenRouter live pricing 2026-07-09: identical prompt/completion prices for base and -pro); they cost more per task by consuming more tokens, not a higher rate. - Context lengths need no new entries: "gpt-5.6-sol" et al. are substrings of their -pro variants and both lookup tables match longest-key-first (verified: sol-pro -> 1.05M direct / 272K codex). - model_switch sort: -pro variants parse as suffix "sol-pro" (rank 1), so /model gpt still defaults to base sol — pinned by test. - Not added to DEFAULT_CODEX_MODELS: only confirmed routable via API/ OpenRouter so far; codex live discovery will surface them if ChatGPT exposes them, same policy as other unconfirmed codex slugs. Tests: invariant tests extended (pro aliases share base entries, base sol outranks sol-pro); 191 targeted tests pass.
83 lines
3.6 KiB
Python
83 lines
3.6 KiB
Python
"""Behavior contracts for the GPT-5.6 (Sol/Terra/Luna) registration.
|
|
|
|
Invariant tests only — no list snapshots (per the no-change-detector-tests
|
|
policy). These pin the two behaviors that would silently regress:
|
|
|
|
1. Version/tier sorting: the flagship Sol must outrank Terra/Luna and the
|
|
whole 5.6 series must outrank 5.5, so `/model gpt` resolves to the
|
|
flagship rather than an alphabetical-first cheap tier.
|
|
2. Pricing reachability: the ("openai", <model>) official-docs pricing keys
|
|
must be reachable from BOTH the bare "openai" provider and the
|
|
"openai-api" picker slug (resolve_billing_route normalizes the latter).
|
|
"""
|
|
|
|
from decimal import Decimal
|
|
|
|
from agent.usage_pricing import (
|
|
_OFFICIAL_DOCS_PRICING,
|
|
_lookup_official_docs_pricing,
|
|
resolve_billing_route,
|
|
)
|
|
from hermes_cli.model_switch import _model_sort_key
|
|
|
|
|
|
class TestGpt56SortInvariants:
|
|
def test_sol_outranks_terra_and_luna(self):
|
|
models = ["gpt-5.6-luna", "gpt-5.6-terra", "gpt-5.6-sol"]
|
|
models.sort(key=lambda m: _model_sort_key(m, "gpt"))
|
|
assert models[0] == "gpt-5.6-sol"
|
|
|
|
def test_56_series_outranks_55(self):
|
|
models = ["gpt-5.5", "gpt-5.5-pro", "gpt-5.6-sol"]
|
|
models.sort(key=lambda m: _model_sort_key(m, "gpt"))
|
|
assert models[0] == "gpt-5.6-sol"
|
|
|
|
def test_aggregator_prefix_form(self):
|
|
models = ["openai/gpt-5.5-pro", "openai/gpt-5.6-sol"]
|
|
models.sort(key=lambda m: _model_sort_key(m, "openai/gpt"))
|
|
assert models[0] == "openai/gpt-5.6-sol"
|
|
|
|
|
|
def test_base_sol_outranks_sol_pro_for_alias_default(self):
|
|
# "-pro" high-effort variants parse as suffix "sol-pro" (rank 1), so
|
|
# `/model gpt` defaults to base Sol rather than the high-effort mode.
|
|
models = ["gpt-5.6-sol-pro", "gpt-5.6-sol"]
|
|
models.sort(key=lambda m: _model_sort_key(m, "gpt"))
|
|
assert models[0] == "gpt-5.6-sol"
|
|
|
|
|
|
class TestGpt56PricingRoute:
|
|
def test_official_pricing_reachable_from_openai(self):
|
|
route = resolve_billing_route("gpt-5.6-sol", provider="openai")
|
|
entry = _lookup_official_docs_pricing(route)
|
|
assert entry is not None
|
|
assert entry.input_cost_per_million == Decimal("5.00")
|
|
|
|
def test_official_pricing_reachable_from_openai_api_slug(self):
|
|
# "openai-api" is the picker slug for direct api.openai.com and must
|
|
# normalize to the "openai" pricing key space.
|
|
route = resolve_billing_route("gpt-5.6-sol", provider="openai-api")
|
|
assert route.provider == "openai"
|
|
entry = _lookup_official_docs_pricing(route)
|
|
assert entry is not None
|
|
assert entry.input_cost_per_million == Decimal("5.00")
|
|
|
|
def test_cache_write_is_1_25x_input_for_56_series(self):
|
|
for slug in ("gpt-5.6-sol", "gpt-5.6-terra", "gpt-5.6-luna"):
|
|
entry = _OFFICIAL_DOCS_PRICING[("openai", slug)]
|
|
assert entry.input_cost_per_million is not None, slug
|
|
assert entry.cache_write_cost_per_million == (
|
|
entry.input_cost_per_million * Decimal("1.25")
|
|
), slug
|
|
assert entry.cache_read_cost_per_million == (
|
|
entry.input_cost_per_million * Decimal("0.10")
|
|
), slug
|
|
|
|
def test_pro_variants_alias_to_base_tier_pricing(self):
|
|
# -pro high-effort modes bill at the same per-token rates as their
|
|
# base tiers; the snapshot aliases them rather than duplicating rows.
|
|
for base in ("gpt-5.6-sol", "gpt-5.6-terra", "gpt-5.6-luna"):
|
|
assert (
|
|
_OFFICIAL_DOCS_PRICING[("openai", f"{base}-pro")]
|
|
is _OFFICIAL_DOCS_PRICING[("openai", base)]
|
|
), base
|